On July 16, 2013 12:00:36 PM Gabriel Littman wrote: > Please correct me if I'm wrong, but I think the right answer is doing a GET > first so you have a vector clock that is after the delete. Then you should > be able to be sure your new write wins in any sibling resolution. > > Gabe > I would assume this to be the case. I've even made sure my delete requests all have vclocks attached (by doing a GET first). However I still see missing keys every so often.
By making each key unique across runs of the test suite seems to solve things, but I don't like that as a solution. And I'm not sure if this is an actual bug or not. If this behaviour is considered a bug, I can try making a minimal reproducible test case and file a bug using that. -- Matthew
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ riak-users mailing list riak-users@lists.basho.com http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
